Authors: Shoorangiz Shams Shamsabad Farahani Mohammad Reza JahedMotlagh Mohammad Ali Nekoui Seyed Vahid Azhari
Publish Date: 2013/05/14
Volume: 73, Issue: 4, Pages: 2291-2311
Abstract
Due to the recent advances in communication networks congestion control can be considered as of paramount importance and has spurred a significant amount of research In this paper the problem of congestion control in the presence of channel uncertainty is addressed in wireless sensor networks WSNsSince WSNs are normally set up in adverse conditions robustness against environmental strains and also the ability to autonomously recover from error situations is mostly demanded Based on nonlinear dynamics point of view it is essential to consider delay in the analysis of general congestion control schemes This paper presents a novel robust decentralized adaptive nonquadratic congestion control scheme RDANQCC which is robust to queue length changes and the consequent delay changes RDANQCC includes an adaptive backoff interval selection that works in concert with a robust controller Using queue utilization and channel estimation algorithm congestion can be detected and a suitable rate is enforced by adaptive backoff interval selection An iterative linear matrix inequality ILMI is utilized to solve our control problem Unlike previous schemes varying desired queue length is considered in this paper and two cases are proposed for fixed and varying desired queue length and it is shown that in the latter our controller renders better performance compared with the former The convergence analysis of RDANQCC is demonstrated by the use of nonquadratic Lyapunovbased approach Extended simulation results demonstrate that RDANQCC renders better performance in comparison with other schemes available in the literature
Keywords: